Silencing of Glycopeptide Resistance in Enterococcus faecalis BM4405 by Novobiocin

Abstract

Enterococcus faecalis BM4405-1, a susceptible derivative of the VanE-type vancomycin-resistant E. faecalis strain BM4405, was obtained after growth in the presence of novobiocin, an inhibitor of the GyrB subunit of DNA gyrase. In contrast to findings for BM4405, UDP-MurNAc-l-Ala-γ-d-Glu-l-Lys-d-Ala-d-Ala (pentapeptide[d-Ala]) was the only peptidoglycan precursor found in BM4405-1, and no VanXY_E d,d-peptidase or VanT serine racemase activities were detected in that strain, even after induction by subinhibitory concentrations of vancomycin. Sequencing of the vanE operon of BM4405-1 revealed two mutations leading to substitutions in VanE (D200N) and in the C-terminal amino acid of VanR_E (Y225F). Cloning of the vanE, vanXY_E, and vanT_E genes of BM4405-1 into the susceptible E. faecalis strain JH2-2 conferred resistance to vancomycin, indicating that the mutation in vanE was not responsible for susceptibility. Transcriptional analysis of the vanE operon in BM4405 by quantitative reverse transcription-PCR indicated that novobiocin did not affect the expression level of the vanE operon. Sequencing of the gyrB gene of BM4405-1 revealed a mutation responsible for substitution of a residue (K337Y) required for ATPase activity and thus implicated in DNA supercoiling. Cloning of the gyrB gene of BM4405 restored vancomycin resistance to BM4405-1. Taken together, these data suggest that alteration of DNA supercoiling following a mutation in GyrB was responsible for lack of expression of the vanE operon and thus for vancomycin susceptibility in BM4405-1.

In gram-positive bacteria, glycopeptides inhibit the last steps of peptidoglycan synthesis by binding to the C-terminal d-alanyl-d-alanine (d-Ala-d-Ala) dipeptide of peptidoglycan precursors, preventing their incorporation into the cell wall (22). Five types of acquired glycopeptide resistance, mediated by VanA, -B, -D, -E, and -G, and one type of intrinsic resistance, mediated by VanC, have been described for enterococci. Resistance to glycopeptides is due to production of modified peptidoglycan precursors, ending in d-alanyl-d-lactate (d-Ala-d-Lac) for VanA, VanB, and VanD or in d-alanyl-d-serine (d-Ala-d-Ser) for VanC, VanE, and VanG, which exhibit reduced affinity for vancomycin (7). VanE-type resistance to low levels of vancomycin in Enterococcus faecalis BM4405 (13) is due to acquisition in the chromosome of the vanE operon, which includes five genes (2). The vanE gene encodes a ligase that synthesizes the dipeptide d-Ala-d-Ser; vanXY_E encodes a bifunctional d,d-peptidase that possesses d,d-dipeptidase and d,d-carboxypeptidase activities, allowing hydrolysis of precursors ending in d-Ala; and vanT_E encodes a serine racemase that provides d-Ser for the resistance pathway. In addition to these three genes, which are sufficient to confer vancomycin resistance, two genes, vanR_E and vanS_E, encode a two-component regulatory system that is implicated in the regulation of expression of the resistance genes (2, 8). The five genes are cotranscribed from a P_E promoter located upstream from vanE (2). Although the VanS_E sensor is likely to be inactive due to the presence of a stop codon in the 5′ portion of the gene, expression of vancomycin resistance is inducible in BM4405, suggesting cross talk with another two-component system (2).

In bacteria, DNA supercoiling is essentially controlled by two enzymes: DNA gyrase and topoisomerase IV. Whereas the first enzyme increases supercoiling in an energy-dependent reaction, the second, in the presence of ATP and Mg²⁺, removes supercoiling and relaxes DNA (10). DNA gyrase is an A₂B₂ complex in which the A subunit is composed of an N-terminal domain involved in DNA breakage and reunion and a C-terminal domain involved in DNA-protein interactions and the B subunit consists of an N-terminal domain containing ATPase activity and a C-terminal domain implicated in interactions with the A protein and DNA (18, 24). Coumermycin and novobiocin, which belong to the coumarin family, specifically inhibit DNA gyrase activity (14). The coumarins block the introduction of supercoils into relaxed DNA and relax supercoiled chromosomal DNA by inhibiting the ATPase activity of DNA gyrase (14). Consequently, DNA replication and cell growth are stopped. Intercalating agents that remove negative supercoils and subinhibitory doses of novobiocin can cause plasmid “curing,” that is, a preferential loss of extrachromosomal DNA (19, 26). The N-terminal domain of GyrB consists of a C-terminal and an N-terminal subdomain (24 kDa), the latter containing the coumarin-binding site (17, 29).

Prior to the demonstration that the vanE operon was chromosomally located in E. faecalis BM4405 (2), the strain was subjected to plasmid curing by novobiocin in order to determine the genetic support of this acquired operon. Because susceptible derivatives were repeatedly recovered at high frequencies, the aim of this work was to characterize one of these variants.

MATERIALS AND METHODS

Strains, plasmids, and growth conditions.

The bacterial strains and plasmids used in this study are described in Table 1. E. faecalis BM4405 is resistant to low levels of vancomycin (MIC, 16 μg/ml) (13). E. faecalis BM4405-1, obtained after growth in the presence of 12.5 μg of novobiocin/ml, was susceptible to vancomycin (MIC, 2 μg/ml). E. faecalis JH2-2, used in electrotransformation experiments, is susceptible to glycopeptides and resistant to fusidic acid and rifampin (16). Escherichia coli Top10 (Invitrogen, Groningen, The Netherlands) was used as the host in cloning experiments. Strains were cultured in brain heart infusion (BHI) broth or agar (Difco Laboratories, Detroit, Mich.) at 37°C. Susceptibility to glycopeptides was determined by Etest as described by the manufacturer (AB Biodisk, Solna, Sweden). The MIC of novobiocin was determined by the method of Steers et al. (25) with 10⁵ CFU per spot on BHI agar after 24 h of incubation at 37°C. Susceptibility to novobiocin was also tested by the disk diffusion method on BHI agar.

TABLE 1.

Strains and plasmids

Strain or plasmid	Relevant characteristicsa	Source or reference
Strains
E. coli Top10	F−[lacIq Tn10(Tcr)] mcrAΔ(mrr-hsdRMS mcrBC) φ80 lacZΔM15ΔlacX74 recA araD139 Δ(ara-leu)7697 galU galK rpsL (Strr) endA1 nupG	Invitrogen
E. faecalis BM4405	Vmr (VanE type)	13
E. faecalis BM4405-1	Vms derivative of BM4405	This work
E. faecalis JH2-2	Fusr Rifr	16
Plasmids
PCR2.1	Apr KmrlacZα oriRColE1	Invitrogen
pCRBlunt	Kmr ZeorlacZα oriRColE1 ccdB	Invitrogen
pAT29	SprlacZα oriRpAMβ1 oriRpUC oriTRK2	27
pAT669	4.4-kb fragment (600 bp, vanEXYETE) of BM4405-1 cloned into pAT29	This work
pAT670	2.4-kb PCR fragment (gyrB) of BM4405 cloned into pAT29	This work

^aAp^r, ampicillin resistance; Fus^r, fusidic acid resistance; Rif^r, rifampin resistance; Sp^r, spectinomycin resistance; Str^r, streptomycin resistance; Tc^r, tetracycline resistance; Vm^r, vancomycin resistance; Zeo^r, zeocin resistance.

Plasmid-curing experiments.

Dilutions of an E. faecalis BM4405 culture in BHI broth were flooded onto tryptic soy agar (TSA) containing 12.5 μg of novobiocin/ml, and the plates were incubated overnight at 37°C. Colonies were screened for vancomycin susceptibility by replica plating on TSA without and with vancomycin (4 μg/ml).

Plasmid and strain construction. (i) Plasmid pAT669.

A fragment of BM4405-1 total DNA encompassing the vanE, vanXY_E, and vanT_E genes and 600 bp upstream from vanE was amplified by using primers E65 and TE6 (Table 2), containing a SacI and an XbaI site, respectively. The PCR product was digested with SacI and XbaI and cloned into pAT29 (27). DNA from recombinant plasmid pAT669 (see Fig. 2C) was introduced into E. faecalis JH2-2 by electrotransformation, and transformants were selected with spectinomycin (60 μg/ml).

TABLE 2.

Oligonucleotides used in this study

Primer	Sequence	Positiona
PCR mapping
E6	5′ GTTTAGCAGTGGATTTGG	1329-1346
E8	5′ CCAAATCCACTGCTAAAC	1346-1329
E15	5′ CAGAAGCTGAGCTAGT	2576-2561
E37	5′ GGATCACCGAAGAAGGT	119-135
E65	5′ GCGCGAGCTCACAGATCAGGAAATCGGAb	NAc
SE2	5′ GGAGTTCTTAAGTCATGT	4908-4891
TE3	5′ GGTTAGGTACAGAGT	3688-3702
TE6	5′ GCGCATGCTCTAGAAGCCATTAAACATCCTd	3785-3770
VRM1	5′ GTTATGGCGCATATTGCT	4365-4382
VRM2	5′ TAACTCTTCTGACGATATT	4319-4301
Cloning of BM4405 gyrB
gyrB6	5′ GCGCGAGCTCGAAACGATTGAAGGCAAAGb	NA
gyrB8	5′ GCTCTAGATTGTGCCATCCGCACCd	NA

^aNucleotide numbering begins at the first base of the vanE gene.

^bThe SacI site is underlined.

^cNA, not applicable.

^dThe XbaI site is underlined.

FIG. 2.

Schematic representation of the vanE gene cluster in vancomycin-susceptible derivative BM4405-1. (A) Open arrows represent coding sequences and indicate the direction of transcription. The asterisk indicates the stop codon in vanS_E. Vertical arrows indicate mutations leading to amino acid substitutions. (B) PCR mapping of the vanE operon. Arrowheads indicate positions and orientations of primers. Horizontal lines represent the PCR products. (C) The insert in recombinant plasmid pAT669 is represented by a solid line (the vector is given in parentheses).

(ii) Plasmid pAT670.

Based on the sequence of the gene for the B subunit of DNA gyrase (GenBank accession number AB059405), the gyrB gene from BM4405 with its ribosome binding site was amplified by PCR with total DNA as a template and oligodeoxynucleotides gyrB6 and gyrB8 (Table 2), containing a SacI and an XbaI site, respectively. The amplification product was digested with SacI and XbaI, and the insert was ligated in pAT29, yielding plasmid pAT670. The recombinant plasmid was introduced into E. faecalis BM4405-1 by electrotransformation, with selection on spectinomycin (200 μg/ml).

Pulsed-field gel electrophoresis.

Genomic DNA embedded in agarose plugs was digested with SmaI overnight at 25°C. Fragments were separated on a 0.8% agarose gel with a CHEF (contour-clamped homogenous electric field)-DRIII system (Bio-Rad Laboratories, Hercules, Calif.) under the following conditions: total migration, 24 h; initial pulse, 60 s; final pulse, 120 s; voltage, 6 V/cm; included angle, 120°; temperature, 14°C. DNA was transferred to Hybond N⁺ membranes (Amersham Pharmacia Biotech, Freiburg, Germany) and fixed under UV illumination. Plasmid pAT663 DNA (vanE′) (13) labeled with [α-³²P]dCTP (Amersham Pharmacia Biotech) by nick translation was used as a probe. Southern blot experiments were carried out under stringent conditions.

PCR and DNA sequencing.

The promoter region and the vanE gene cluster (see Fig. 2A) were amplified by using BM4405-1 DNA as a template and primer pairs specific for the vanE operon as described elsewhere (2). The gyrB genes of BM4405 and BM4405-1 were amplified by using primer pair gyrB6-gyrB8 (Table 2). DNA amplification was carried out in a GeneAmp PCR system 2400 thermal cycler (Perkin-Elmer Cetus, Norwalk, Conn.). The PCR products were ligated into pCR2.1 and transformed into E. coli Top10 (Invitrogen). Plasmid DNA was extracted with the commercial Wizard Plus Minipreps DNA purification system (Promega, Madison, Wis.) and labeled with a Dye-Labeled ddNTP Terminator Cycle Sequencing kit (Beckman, Fullerton. Calif.), and the samples were sequenced and analyzed with a CEQ 2000 automated sequencer (Beckman).

Analysis of peptidoglycan precursors.

Peptidoglycan precursors were extracted and analyzed as described elsewhere (20). Strains were grown in BHI broth overnight at 37°C in the absence or in the presence of vancomycin (1 μg/ml for BM4405-1 and 4 μg/ml for JH2-2/pAT669) with gentle agitation to an optical density at 600 nm (OD₆₀₀) of 1.0 (mid-exponential phase). Ramoplanin was added to a concentration of 3 μg/ml, and incubation was continued for 15 min. Bacteria were harvested, and the cytoplasmic precursors were extracted with 8% trichloroacetic acid (15 min at 4°C), desalted, and analyzed by high-performance liquid chromatography. Results were expressed as the percentages of total late peptidoglycan precursors represented by UDP-MurNAc-l-Ala-γ-d-Glu-l-Lys-d-Ala (UDP-MurNAc-tetrapeptide), UDP-MurNAc-l-Ala-γ-d-Glu-l-Lys-d-Ala-d-Ala (UDP-MurNAc-pentapeptide[d-Ala]), and UDP-MurNAc-l-Ala-γ-d-Glu-l-Lys-d-Ala-d-Ser (UDP-MurNAc-pentapeptide[d-Ser]) that were determined from the integrated peak areas.

d,d-Dipeptidase, d,d-carboxypeptidase, and serine racemase activities.

Enzymatic activities were assayed in BM4405-1 as described previously (5). Bacteria grown in BHI broth overnight at 37°C in the absence or presence (1 μg/ml) of vancomycin were lysed by treatment with lysozyme (2 μg/ml) at 37°C, followed by sonication, and the membrane fraction was pelleted (at 100,000 × g for 45 min). Activities were measured in the supernatant (S100) and in the resuspended pellet (C100) fraction. Cytoplasmic and membrane fractions (15 μl) were incubated in Bis-Tris propane (150 mM; pH 7.5) with 10 mM d-Ala-d-Ala (VanX activity), 5 mM pentapeptide (VanY activity), or 10 mM l-Ser (VanT activity) for 30 min at 37°C. The d-Ala released from d-Ala-d-Ala by d,d-dipeptidase or from the pentapeptide (l-Ala-γ-d-Glu-l-Lys-d-Ala-d-Ala) by d,d-carboxypeptidase and the d-amino acids produced by racemase activity with d-Ser as a standard were detected by using d-amino acid oxidase with o-dianisidine as a chromogen (20). Specific activities were defined as the number of nanomoles of product formed at 37°C per minute per milligram of protein contained in the extracts.

RNA techniques. (i) Extraction of total RNA.

Total RNA was extracted with the commercially available TOTALLY RNA kit (Ambion, Austin, Tex.) as follows. E. faecalis BM4405-1 was grown to an OD₆₀₀ of 0.7. Cells were harvested by centrifugation, resuspended in TE (Tris-HCl, 10 mM; EDTA, 1 mM), and incubated in the presence of lysozyme (1 mg/ml) for 5 min at room temperature. After centrifugation (for 5 min at 15,000 × g), supernatants were extracted in a first step with a mixture containing phenol, chloroform, and isoamyl acohol, followed by a second step with a mixture composed of sodium acetate, acid phenol, and chloroform. Total RNA was precipitated by addition of 1 ml of isopropanol, and the RNA pellets were resuspended in water-diethyl pyrocarbonate plus EDTA. RNA quality was assessed by agarose gel electrophoresis. RNA concentrations were determined by measuring the absorbance at 260 nm.

(ii) Northern blot analysis.

Northern blot experiments were performed as described previously (2). Fragments obtained by PCR using total DNA from BM4405 as a template and primers E37-E38 (vanE), XYE1-XYE2 (vanXY_E), E9-E15 (vanT_E), R2-R6 (vanR_E), and S1-E41 (vanS_E) (2) were labeled with [α-³²P]dCTP (3,000 Ci/mmol; Amersham Pharmacia Biotech) by using the Megaprime DNA labeling system (Amersham Pharmacia Biotech). Hybridization and washes were performed as described previously (2). The sizes of the transcripts were determined by using RNA molecular weight marker I (Boehringer, Mannheim, Germany).

(iii) RT-PCR experiments.

Total-RNA samples were prepared as described previously (2). Reverse transcription (RT) was carried out as described previously (2) using 50 pmol of primer vanE2 (5′-GTCGATTCTCGCTAATCC). The DNA products were amplified by PCR in an 80-μl reaction volume containing the previous 20-μl samples of the RT product, 50 pmol each of primers vanE1 (5′-TGTGGTATCGGAGCTGCAG) and vanE2, 1× enzyme buffer (QBIOgene, Montreal, Canada), and 2 U of Taq DNA polymerase (QBIOgene). PCR was performed with the following thermal cycling profile: 3 min at 94°C; 18 cycles of amplification consisting of 1 min at 94°C, 1 min at 52°C, and 1 min at 72°C; and a final extension for 7 min at 72°C. RT-PCR products were loaded onto a 1.2% agarose gel and quantified by using Quantity One software (Bio-Rad).

(iv) Primer extension analysis.

The putative transcriptional start site for the vanE operon of BM4405 grown in the presence of novobiocin alone (4 μg/ml) or of novobiocin and vancomycin (each at 4 μg/ml) was located by primer extension using the synthetic oligodeoxynucleotide PE1 (5′-CCAATGACCTTCTTCGGTGATCC) as described elsewhere (2).

RESULTS AND DISCUSSION

Characterization of strain BM4405-1.

The E. faecalis VanE-type clinical isolate BM4405 is resistant to 16 μg of vancomycin/ml (13). After growth of the strain in the presence of novobiocin (12.5 μg/ml), 7 to 12% of the surviving cells were susceptible to vancomycin (MIC, ≤2 μg/ml), and one derivative, BM4405-1, was selected for further studies (Table 1). Strains BM4405 and BM4405-1 were compared by pulsed-field gel electrophoresis (Fig. 1). Their SmaI patterns were indistinguishable, and the same-size fragment (ca. 600 kb) from both strains hybridized with a vanE-specific probe (Fig. 1).

FIG. 1.

Analysis of SmaI-digested total DNA from E. faecalis isolates by pulsed-field gel electrophoresis (left) and by Southern hybridization (right) with a vanE-specific probe. Lanes: 1, molecular weight marker (bacteriophage λ DNA); 2, BM4405 (resistant to vancomycin); 3, JH2-2; 4 BM4405-1 (susceptible to vancomycin); 5, molecular weight marker (Saccharomyces cerevisiae chromosomes).

By disk diffusion, BM4405-1 exhibited a reproducibly slightly reduced susceptibility to novobiocin (inhibition zone diameter, 13 mm) relative to that of BM4405 (inhibition zone diameter, 15 mm). In addition, the growth of BM4405 was slower than that of BM4405-1 in the presence of noviobiocin (12 μg/ml) in the culture medium. However, this did not translate into a difference in the novobiocin MIC (16 μg/ml) as determined by agar dilution using an arithmetic progression.

The organization of the vanE operon in BM4405-1 was determined by PCR mapping with primer pairs specific for every gene of the operon in BM4405 (Fig. 2B; Table 2). Amplified fragments of the expected sizes were obtained, indicating that all the genes were present in BM4405-1 and were in the same order as in strain BM4405 (2). Determination of the sequence of the operon detected a silent mutation in vanXY_E (T1376A) and two mutations leading to amino acid substitutions. The first, in vanE (G597A), resulted in the replacement of aspartate by asparagine (D200N). The second mutation, in the 3′-terminal end of the vanR_E gene, resulted in the replacement of the terminal tyrosine at position 225 by phenylalanine (Y225F). The stop codon previously found in vanS_E of BM4405 (2) was still present in BM4405-1.

Peptidoglycan precursors.

To analyze the cytoplasmic peptidoglycan precursors, strain BM4405-1, grown in the absence or presence of a subinhibitory concentration of vancomycin (1 μg/ml), was incubated in the presence of ramoplanin to inhibit cell wall synthesis immediately after synthesis of the precursors (20). The results obtained indicated that UDP-MurNAc-pentapeptide was the only late precursor synthesized by noninduced or induced BM4405-1 (Table 3), as in E. faecalis BM4405 grown in the absence of vancomycin (13).

TABLE 3.

Glycopeptide MICs and nature of the peptidoglycan precursors in E. faecalis strains

E. faecalis strain	MIC (μg/ml) of vancomycin	% of total precursorsa represented by:
		Tetra	Penta[d-Ser]	Penta[d-Ala]
BM4405	16	2	8	90
BM4405 (Vm 4)b	16	10	90	0
BM4405-1	2	1	0	99
JH2-2	2	NDc	ND	ND
JH2/pAT669	8	16.5	17	66.5
JH2-2/pAT669 (Vm 1)d	8	26	45	29
BM4405-1/pAT670	8	10	35	55
BM4405-1/pAT670 (Vm 4)b	8	10	70	20

^aTetra, UDP-MurNAc-l-Ala-γ-d-Glu-l-Lys-d-Ala; Penta-d-Ser, UDP-MurNAc-l- Ala-γ-d-Glu-l-Lys-d-Ala-d-Ser; Penta-d-Ala, UDP-MurNAc-l-Ala-γ-d-Glu-l- Lys-d-Ala-d-Ala.

^bMIC after induction with 4 μg of vancomycin/ml.

^cND, not determined.

^dMIC after induction with 1 μg of vancomycin/ml.

d,d-Dipeptidase, d,d-carboxypeptidase, and serine racemase activities.

d,d-Dipeptidase (VanX) and d,d-carboxypeptidase (VanY), respectively, hydrolyze d-Ala-d-Ala and remove the terminal d-Ala residue of precursors ending in acyl-d-Ala-d-Ala (6). In VanC, VanE, and VanG, these two activities are catalyzed by a bifunctional VanXY-type enzyme (12, 23). Enzyme activities were determined on the S100 and C100 fractions of extracts from BM4405-1 cells grown without or with vancomycin (1 μg/ml). Whereas weak d,d-peptidase (VanX and VanY) activities were found in the soluble extracts of induced BM4405 cells (13), no d,d-peptidase activities were detected in the soluble or insoluble fractions from induced or noninduced BM4405-1 (data not shown).

Racemization of l-Ser to d-Ser by a membrane-bound VanT serine racemase provides d-Ser required for peptidoglycan synthesis in enterococci expressing the VanC, VanE, or VanG type of resistance (2-4, 12). Although high vancomycin-inducible serine racemase activity is produced by BM4405 (13), no VanT activity was detected in membrane extracts from strain BM4405-1, even after growth with a subinhibitory concentration (1 μg/ml) of vancomycin. Taken together, these results confirm that the vanE operon is not expressed in BM4405-1.

Functional analysis of the vanE operon of BM4405-1.

Plasmid pAT667, with a 4.4-kb insert that includes the vanEXY_ET_E genes of BM4405 together with 600 bp upstream from vanE, confers inducible vancomycin resistance on E. faecalis JH2-2 (2). To test if the mutation in the vanE gene of BM4405-1 was responsible for the susceptibility of the host, the three resistance genes and 600 bp upstream from vanE were cloned into pAT29, yielding plasmid pAT669 (600 bp, vanEXY_ET_E) (Table 1), which was introduced into E. faecalis JH2-2 by electrotransformation. The vancomycin MIC for the transformant (8 μg/ml) differed significantly from that for JH2-2 (2 μg/ml) (Table 3). The nature of the peptidoglycan precursors of JH2-2/pAT669 was determined after growth in the absence or presence of vancomycin (4 μg/ml). In the absence of vancomycin, UDP-MurNAc-pentapeptide[d-Ala] was mainly synthesized, whereas in the presence of vancomycin, UDP-MurNAc-pentapeptide[d-Ser] was the main precursor produced (Table 3). Susceptibility to teicoplanin remained unchanged even after induction with vancomycin. These results are similar to those obtained with strain JH2-2/pAT667 (2). Since the vanEXY_ET_E genes of BM4405-1 conferred inducible vancomycin resistance when introduced into a susceptible E. faecalis strain, the mutation in the vanE gene of BM4405-1 did not account for significant loss of vancomycin resistance in that strain.

Transcriptional analysis of the vanE operon.

The start codons of the vanXY_E and vanT_E genes overlap the termination codons of vanE and vanXY_E, respectively, suggesting that the three genes are cotranscribed (2). In E. faecalis BM4405, a single transcript of ca. 5,800 bases hybridizes with all the genes in the operon, including vanR_E and vanS_E (2). Total RNA from BM4405-1 was analyzed by Northern hybridization with probes internal to every gene in the operon (2), but no transcript was detected (Fig. 3). No transcripts were detected after induction (vancomycin, 1 μg/ml) of BM4405-1 (data not shown).

FIG. 3.

Transcription analysis of the vanE gene cluster by Northern hybridization. Total RNA was hybridized with vanE (A), vanXY_E (B), vanT_E (C), vanR_E (D), or vanS_E (E) probes from induced (lanes 1) and noninduced (lanes 2) BM4405 (resistant to vancomycin) and from noninduced BM4405-1 (susceptible to vancomycin) (lanes 3). The sizes of the transcripts were determined by reference to RNA molecular weight marker I (Boehringer) (not shown). b, bases.

Primer extension was performed to locate the putative transcriptional start site when BM4405 was grown in the absence or presence of novobiocin alone (4 μg/ml) or novobiocin plus vancomycin (4 μg/ml each) (Fig. 4). By using primer PE1, complementary to the 5′ end of the vanE gene, the transcriptional start site was located, under all growth conditions, 291 bp upstream from the ATG of the vanE gene (Fig. 4), as determined previously for that strain (2). This observation suggests that novobiocin did not directly affect the transcriptional start of the vanE operon.

FIG. 4.

Identification of the transcriptional start site for the vanE, vanXY_E, vanT_E, vanR_E, and vanS_E genes by primer extension analysis. (Left) Primer elongation product obtained with oligodeoxynucleotide PE1 and 50 μg of total RNA from BM4405 grown in the absence (lane 1) or in the presence (4 μg/ml) (lane 2) of novobiocin or in the presence of vancomycin plus novobiocin (each at 4 μg/ml) (lane 3). Lanes T, G, C, and A, results of sequencing reactions performed with the same primer. (Right) Sequence from nucleotide positions −353 to +141 (numbering from the A of the ATG start codon of vanE, negative in the 3′-to-5′ direction and positive in the 5′-to-3′ direction). The +1 transcriptional start site for the vanE, vanXY_E, vanT_E, vanR_E, and vanS_E mRNA in BM4405 and the −35 and −10 promoter sequences located upstream are boldfaced. The ATG start codon of vanE is indicated by an arrow, and the ribosome binding site (RBS) is boldfaced and underlined.

Quantitative RT-PCR was used to determine the level of expression of the vanE operon in BM4405 and BM4405-1 cultivated under various conditions. The strains were grown without antibiotic until an OD₆₀₀ of 0.2 was reached. The cultures were then divided in two subcultures called A and B. Vancomycin (4 μg/ml for BM4405 and 1 μg/ml for BM4405-1) was added to subculture B, and both cultures were incubated to an OD₆₀₀ of 0.7. Subcultures A and B were then divided again into subcultures A1 and A2 and subcultures B1 and B2. Novobiocin (4 μg/ml) was added to subcultures A2 and B2. All four cultures were incubated for 1 h, and samples were taken for RT-PCR. Quantification of the amplification products indicated that the levels of expression of the vanE gene in BM4405 were similar in subcultures A1 and A2 and also in subcultures B1 and B2 (data not shown), indicating that novobiocin had no influence on the level of expression of the vanE operon. Further addition of vancomycin to subculture B resulted, as expected, in an increase (of ca. 30%) in vanE expression in subcultures B1 and B2 relative to that in subcultures A1 and A2 (data not shown). These results confirm that (i) novobiocin did not interfere directly with the expression of the vanE operon and (ii) vancomycin induced the expression of the vanE operon, even in the presence of novobiocin.

No RT-PCR products were obtained with BM4405-1, even after induction with a subinhibitory concentration of vancomycin (1 μg/ml), confirming that the vanE operon was not expressed in the susceptible strain.

Analysis of the gyrB gene.

We tested the possibility that the gyrB gene in BM4405-1 had suffered mutations that could be responsible for loss of expression of the vanE operon. The sequences of the amplification products obtained by using E. faecalis gyrB-specific primers (Table 2) and total DNA from BM4405 and BM4405-1 as templates were determined. Mutation A1009C was found in the gyrB gene of BM4405-1, leading to replacement of the lysine at position 337 by tyrosine (K337Y). Lysine 337 is known to be involved in ATP binding to the GyrB subunit (15). The gyrB gene of the parental strain, BM4405, was amplified by PCR and cloned into pAT29, and the recombinant plasmid, pAT670 (Table 3), was introduced by electrotransformation into BM4405-1. Transformant BM4405-1/pAT670 was resistant to vancomycin (MIC, 8 μg/ml), indicating that lack of expression of the vanE operon of the host was, at least in part, associated with a defect in DNA gyrase activity due to the mutation in the gyrB gene. This result was confirmed by analysis of the peptidoglycan precursors from BM4405-1/pAT670 performed after growth without or with (4 μg/ml) vancomycin. In the absence of vancomycin, UDP-MurNAc-pentapeptide[Ala] (55%), UDP-MurNAc-pentapeptide[Ser] (35%), and UDP-MurNAc-tetrapeptide (10%) were synthesized, whereas in the presence of vancomycin, UDP-MurNAc-pentapeptide[Ser] represented 70% of the peptidoglycan precursors (Table 3). Determination of the nucleotide sequences of the gyrB genes of five other vancomycin-susceptible BM4405 derivatives obtained independently after treatment with novobiocin revealed a mutation in each of them. Four of these mutations led to an amino acid substitution in the N-terminal subdomain of GyrB (E27Q, M35T, D181G, and G345T), and one led to an amino acid substitution in the C-terminal subdomain (R404H). The consequences of these mutations for the activity of GyrB remain unknown, since they have not been reported previously.

Effects of novobiocin on three other VanE-type E. faecalis strains.

Three distinct VanE-type E. faecalis strains (vancomycin MICs, 6 to 32 μg/ml) (1) were subjected to novobiocin treatment. No derivatives susceptible to vancomycin were obtained in two independent experiments. It seems, therefore, that the effect of novobiocin on expression of the vanE operon is strain dependent. Transcription initiation at numerous bacterial promoters is strongly dependent on supercoiling of the DNA template (9, 21, 28). Primer extension was performed to locate the putative transcriptional start site of the vanE operon in these strains (1). The putative P_E promoter, located upstream of vanE, was identical in the three strains but differed from that in BM4405. The +1 transcriptional start site was located 25 and 291 bp upstream from the start codon in the three strains and in BM4405, respectively (1, 2). Promoters whose expression depends on a transcriptional activator do not possess a typical −35 region (11). Thus, the spacing between the −35 and −10 hexamers, which is crucial for promoter activity (28), could not be determined. The lack of a novobiocin effect on the expression of the vanE operon in the three strains could be due to the fact that their P_E promoter differs from that of BM4405. In two of these strains expression of the vanE operon was inducible, whereas it was constitutive in the isolate that had a mutation in the vanS_E gene (1). Taken together, these results are consistent with the previous proposal (2) that another two-component regulatory system is responsible for expression of the vanE operon in BM4405 by cross talk.

In conclusion, we have shown that novobiocin treatment of the VanE-type strain BM4405 is indirectly responsible for inactivation of the vanE operon. Mutation of a critical residue (K337Y) in GyrB involved in ATP hydrolysis is likely to be responsible for altered expression of the vanE gene cluster. It appears, therefore, that the level of DNA supercoiling modulates the expression of the vanE operon at the transcriptional level. We have previously proposed that, since the VanS_E sensor of BM4405 is not functional, inducible expression of the vanE operon is likely to be due to cross talk with another two-component regulatory system of the cell (2). Whether alteration of DNA supercoiling directly affects the P_E promoter or the activity of the putative other two-component system remains to be determined.